Recent observations indicate that the expression of Ure2p occurs in two manners. The predominant form is via cap-dependent initiation of translation. However, under conditions of "stress", a shorter form of Ure2p is expressed via an IRES that is contained within the coding region of the Ure2 mRNA (corresponding to the Nterminal region of the protein). An apparently unrelated finding has been the identification of the yeast homolog to the human eIF2A gene which is non-essential for yeast growth. However, when the expression of the smaller form of the Ure2p protein (via a reporter system fused to lacZ) was monitored in the eIF2A deletion strain, expression oflacZ was increased 6 to 10-fold. In the simplest of terms, eIF2A would appear to be a suppressor of the IRES function found in the Ure2 mRNA. We propose the following 4 questions to more fully evaluate these findings: 1. What are the cis-acting sequences in Ure2p mRNA that are required to allow internal initiation? 2. What other yeast mRNAs are initiated via internal initiation? 3. How is control of internal initiation regulated, through covalent modification or protein content? 4. What portion of eIF2A is required to suppress internal initiation and with what proteins does it interact in this process? Given the general observation that in many systems, "stress" leads to cap-independent initiation (i.e. internal initiation), we feel that we have discovered an ideal model system in which to study the switch from capdependent to cap-independent translation. It is anticipated that the mechanism defined in the yeast system will have broad applicability and will likely provide insights into mammalian systems where there have been numerous reports on the utilization of cap-independent translation (heat shock, apoptosis, viral infection, etc.)